Recent theoretical work suggests that salt fingers have sufficiently rapid growth rates in much of the thermocline to produce significant vertical fluxes of heat and mass, even an active internal wave field. Using laboratory determinations of the 1/3 power law for the salt flux we have estimated the fingers to be intermittently active on the high-gradient sheets in the thermocline. The flux is found to be comparable to the surface input of salt due to evaporation. This implies a release of energy by the falling out of salt, comparable to the rate of work done by the wind on the mean circulation, and supports the model Stern (1969h). Salt fingers transport heat and mass at different rates, and thus the usual concept of an 'eddy diffusivity' is invalid when salt fingers make a major contribution to the vertical mixing processes.